Definitions have no place in philosophy

Jamal

First, a definition (bear with me):

A definition is a statement that specifies the correct use of a term.

One way of doing this is by identifying necessary and sufficient conditions. But what exactly do I mean by “necessary” and “sufficient”? The SEP article that I’ve just linked to notes that the attempt to define these terms is an elusive quest. Have we reached an impasse then? Not at all: I trust that nobody, once acquainted with examples, really needs any definitions to use the terms—to discuss, let’s say, the necessary and sufficient conditions that allow you to identify what is and what isn’t a work of art.

What’s more, does anyone even need a definition of “definition” to talk about definition? That seems implausible. So why did I define it?

Although I seemed to be starting out by “defining my terms,” in the way that some people in philosophical discussions demand, what I was really doing was explicating a concept that we’re all familiar with, and I was not aiming for comprehensiveness. I was beginning an analysis of a term which we already understand and know how to use; or, to put it differently, I was beginning to describe what we look for in a definition. It may have been a useful exercise, but not because you didn’t already know what a definition is, and not because there’s a likelihood we would end up talking past each other without it.

Does this tell us someting about definitions in general, about what we need them for and when we should provide and demand them? It surely implies that definitions are often not required at all—most often we just need to know how to use a term.

But what of the more contentious terms in philosophy, the big ones such as “knowledge,” “consciousness” and “freedom”?

To address this I'll state my thesis:

A definition of a philosophical concept might be required at the beginning of a discussion only in the case that the term is equivocal.

Note

(For the purposes of this post I’m not regarding stipulative definitions as definitions at all (did you see what I did there?))

While a term may be contentious, it is not thereby equivocal. There are differences in exactly how we understand the terms “freedom” and “consciousness”. You mean the freedom to walk down the street with a gun and I mean the freedom to walk down a street that’s free of lethal weapons. And yet, we’re talking about the same thing, i.e., the term is contentious but not equivocal. The differences here are the stuff of discussion itself—so it might be better to say, not that we understand the term differently or that it has a distinct and separate meaning for each of us, but that it implies different things for us.

We cannot debate how to ensure individual freedom in society if I define your opposing view out of existence. We cannot fruitfully disagree about the nature of consciousness if I attempt to determine the course of the discussion using my favourite necessary and sufficient conditions, conditions that you might not accept. When used in a certain way this is a fallacy, the fallacy of persuasive definition, a mark of sophistry rather than philosophy. Even when it’s not fallacious, it forecloses on certain of the range of possible results.

But all this is really just to say that philosophical discussion, rather than beginning with definitions, should seek to produce them, or at least to approach them by explication—the word “explication” derives from the Latin for “unfold,” and I think this is a good way of thinking about what philosophy does.

Note

It should be noted that the “what is x?” dialogues of Plato end with aporia rather than with a triumphant definition, which might make us wonder if definitions should even be the goal, never mind the starting point.

Wittgenstein has good stuff to say about definitions, but it was Kant who got me interested on this occasion:

In philosophy one must not imitate mathematics by starting from a definition–except perhaps as a mere attempt. For since the definitions are dissections of given concepts, the concepts, though still confused, precede the definitions; and the incomplete exposition precedes the complete one. Thus once we have drawn some characteristics from a still uncompleted dissection, we can infer from them various details in advance, before we have reached the complete exposition, i.e., the definition. In a word, in philosophy the definition, as involving rigorous distinctness, must conclude rather than begin the work. — Kant, Critique of Pure Reason, B758

Concepts are open, their objects change too, and the pertinent conditions in each case are relative to the specifics of a particular topic or discussion. One might say that this is exactly why we need to define our terms, but crucially, definitions always lag behind: we do not know the best way to define our terms when we begin. In this circumstance, our ability to flexibly use terms is superior to our ability to comprehensively define them. Definitions therefore are poor substitutes for a skill, namely the ability to use terms successfully, an ability that does not rest on definitions but on shared meaning.

Kant puts it like this:

[…] I can never be sure that the distinct presentation of a concept given to me (as still confused) has been developed comprehensively, unless I know that it is adequate to the object. However, the object's concept, as it is given, may contain many obscure presentations that we pass over in dissecting the concept, although we always use them in applying it; and hence the comprehensiveness of my concept's dissection is always doubtful … — Kant, Critique of Pure Reason, B756

(My bold)

That is, in our use of a term we carry along some meaning that always gets left out of our definitions.

So what do you think? Is “define your terms!” always or often or ever a legitimate imperative?

Note

By the way, I’m not ashamed of the clickbait title even though I don’t completely endorse it. :razz:

fdrake

one must not imitate mathematics by starting from a definition — Kant, Critique of Pure Reason, B758

I just wanna provide pushback on this linear definition->theorem->proof characterisation of mathematics. As Lakatos highlights in Proofs and Refutations, the concept of "Eulerian polyhedron" was redefined repeatedly over mathematical history to avoid cases which obviously weren't Eulerian polygons. Even in mathematics, a definition is an attempt to explicate a concept, which can be revised if it is insufficient.

The only difference, as I see it, is that creating a definition in mathematics determines what follows from the statement in a relatively formal way - what can be proved given other background assumptions. Whereas in natural language even the means of reasoning vary contextually. I believe there is no such thing as a determinate "step" or "enclosed expression of meaning in natural language, the words always bleed out into the background. Which it seems is something you already wrote in other words.

So what do you think? Is “define your terms!” always or often or ever a legitimate imperative? — Jamal

That said, assuming definition is understood to mean "explication of a concept"; revised after your analysis. A logician's answer to this might be yes. Because you've already characterised how definitions work in natural language. So long as definition taken as a provisional goal for discussion, rather than seen as the step you must take when starting one.

plaque flag

Definitions therefore are poor substitutes for a skill, namely the ability to use terms successfully, an ability that does not rest on definitions but on shared meaning. — Jamal

:up:

Skill looks like the right focus here. Inspired by Brandom and others, I think of applying concepts as a skilled labor, mostly inarticulate cando knowhow, manifesting sensitivity to and respect for the discursive norms we are always already thrown into, which make asking for definitions or after their value possible to begin with.

In my view, it's helpful to emphasize the larger context in which definitions matter. We make and evaluate claims about the world, including what we should do within in it, as part of a community. I claim that it's only because they are used in claims that concepts matter.

In this circumstance, our ability to flexibly use terms is superior to our ability to comprehensively define them. — Jamal

:up:
It's as if we 'are' this skill, and part of that general skill involves the delicate and fragile art of tentative definition. From a 'Hegelian' perspective, concepts are always in flux, slowly drifting. We change the object being clarified (language) as we use it to articulate its own character.

the word “explication” derives from the Latin for “unfold,” and I think this is a good way of thinking about what philosophy does. — Jamal

Beautiful metaphor ! Making It Explicit. If we named global Geistware Shakespeare, we can name the philosophical module Hegel, in honor of someone who made making it explicit explicit to itself. 'Hegel' is that part of spirit (cultural software) which articulates the character of articulation itself.

Concepts are open — Jamal

:up:
The reminds me of discussions of genesis versus structure. That concepts are open make genesis possible. As individuals we can get lucky with a new metaphor which gets adopted becomes relatively literal, hardens like cooling wax. Or we can add to the machinery of metacognition by seeing that maybe the inferential relationships of claims are what make concepts within such claims meaningful, etc.

T Clark

Although I seemed to be starting out by “defining my terms,” in the way that some people in philosophical discussions demand, what I was really doing was explicating a concept that we’re all familiar with, and I was not aiming for comprehensiveness. I was beginning an analysis of a term which we already understand and know how to use; or, to put it differently, I was beginning to describe what we look for in a definition. It may have been a useful exercise, but not because you didn’t already know what a definition is, and not because there’s a likelihood we would end up talking past each other without it. — Jamal

As one of those who might be characterized as demanding definitions, I have some thoughts. The first is one I've expressed here often - many, I would say most, of the frustrating, fruitless discussions we have here on the forum start out with disagreements about the meaning of words and then never make any progress toward actually dealing with any interesting philosophical issues. I don't disagree that discussions where we work out among ourselves what particular terms mean are valuable. I have started a few discussions for that purpose - What does "mysticism" mean; What does "consciousness" mean; What does "real" mean. They were among the more satisfying discussions I've participated in.

On the other hand, I often start discussions about specific issues I want to examine, often something to do with metaphysics. In my OPs I often make it clear exactly what I intend the meaning of specific words are for the purposes of that particular discussion. Then I obnoxiously and legalistically defend that position, sometimes asking moderators to help. I do that because I want to talk about a specific concept or subject and I don't want to argue about what "metaphysics" really means. If I don't make those kinds of requirements, the thread will just turn into an argument about something I'm not interested in.

When I am participating in someone else's discussion, I try to follow their rules. If I am unclear about how they are using a word, I'll ask or I'll say what it means to me. If I don't like the definition of a particular word they are using, I can bug off if it bothers me enough. I often like rough and tumble rhetorical competitions - jokes, insults, and name-calling. But sometimes I just want to get down to work.

Manuel

What's interesting to me is that dictionary definitions barely scratch the surface of the definition of a word. They give a hint as to what the word means, but the application of most words goes way beyond anything given in a dictionary.

Which hints at something which I think is quite profound, and not well understood, because it is obscure. We already know the gist of the meaning of the word, before knowing the actual word. Put in another way, the words add a kind of structure to what we already knew.

As Leibniz said, replying to Locke:

"[Locke]: The crime of killing an old man, not having a name as parricide does, is not taken for a complex idea. [complex idea, roughly meaning=concept]

[Leibniz]: The reason why there is no name for the murder of an old man is that such a name would be of little use... ideas do not depend upon names [words with definitions, in this context] ... If a... writer did invent a name for that crime and devoted a chapter to 'Gerontophony', showing what we owe to the old and how monstrous it is to treat them ungently, he would not thereby be giving us a new idea."

(Emphasis mine.)

I think this applies rather broadly, but there is room for exception.

So, in general, I think that we most of the time, have a decent idea or notion of what we want to communicate. The failure of communication has more to do with the ideas behind the words, than the words themselves. So, I'm inclined to agree that philosophy shouldn't be primarily about definitions, though these can help.

Tom Storm

So what do you think? Is “define your terms!” always or often or ever a legitimate imperative? — Jamal

I will sometimes ask how a person is using a particular term as this is more useful than a 'correct' definition. People who get stuck on specific definitions are often irritating pedants and seem to miss the point.

plaque flag

People who get stuck on specific definitions are often irritating pedants and seem to miss the point. — Tom Storm

:up:

Nickolasgaspar

If people enjoy talking pass each other, then not using definitions is the way to go.

Fooloso4

↪Jamal

I think this overstates the case. The first quoted passage seems to argue against your claim. There are two parts, where we begin and where we aim to end:

In philosophy one must not imitate mathematics by starting from a definition ...

... In a word, in philosophy the definition, as involving rigorous distinctness, must conclude rather than begin the work. — Kant, Critique of Pure Reason, B758

If the definition concludes the work then surely it has a place.

Nickolasgaspar

↪Tom Storm

It depends. Words do not have intrinsic meanings , they have common usages. If someone doesn't use any of the common meanings or his intention is an argument of Ambiguity or we already have a word with that same meaning then the irritating pedant is the guy who doesn't respect the basic rules of communication.

Mww

“…. Thus the criterion of the possibility of a conception (not of its object) is the definition of it, in which the unity of the conception, the truth of all that may be immediately deduced from it, and finally, the completeness of what has been thus deduced, constitute the requisites for the reproduction of the whole conception.…”
(CPR B115)

Note the lower textual location in B only, this in reference to understanding, whereas the other quotes with higher textual locations, refer to pure reason’s dogmatic use, and is found in both editions.

Of course definitions have a place, if only in justifications for a method.

T Clark

People who get stuck on specific definitions are often irritating pedants and seem to miss the point. — Tom Storm

Hey! I resemble that remark.

Tom Storm

↪T Clark

Wasn't thinking of you to be honest. But why not?

Jamal

I just wanna provide pushback on this linear definition->theorem->proof characterisation of mathematics. As Lakatos highlights in Proofs and Refutations, the concept of "Eulerian polyhedron" was redefined repeatedly over mathematical history to avoid cases which obviously weren't Eulerian polygons. Even in mathematics, a definition is an attempt to explicate a concept, which can be revised if it is insufficient. — fdrake

That’s interesting. I hadn’t even thought to question Kant on that. I suppose then that when he says in the same section that “Mathematical definitions never err,” he’s wrong?

But here’s the full passage:

Mathematical definitions can never err. For since the concept is first given through the definition, it contains exactly just what the definition wants us to think through the concept. But although there cannot occur in the concept anything incorrect in content, sometimes–although only rarely–there may still be a defect in the form (the guise) of the concept, viz., as regards its precision. — Kant, Critique of Pure Reason, B759

I wonder if that covers it.

Otherwise I agree.

180 Proof

↪Nickolasgaspar

:up:

Jamal

Skill looks like the right focus here. Inspired by Brandom and others, I think of applying concepts as a skilled labor, mostly inarticulate cando knowhow, manifesting sensitivity to and respect for the discursive norms we are always already thrown into, which make asking for definitions or after their value possible to begin with.

In my view, it's helpful to emphasize the larger context in which definitions matter. We make and evaluate claims about the world, including what we should do within in it, as part of a community. I claim that it's only because they are used in claims that concepts matter. — plaque flag

Seems reasonable.

From a 'Hegelian' perspective, concepts are always in flux, slowly drifting. We change the object being clarified (language) as we use it to articulate its own character. — plaque flag

Yes, I suppose Hegel is the next step here for me. The thing that bothers me about Hegel, and sometimes Adorno too, is the reluctance not only to give definitions—which is justified—but also the reluctance to give examples. Examples are looked down upon by several philosophers, but they’re often what allow me to first get ahold of a concept, and relevantly here, they are part of how we get by without definitions.

Beautiful metaphor ! Making It Explicit. If we named global Geistware Shakespeare, we can name the philosophical module Hegel, in honor of someone who made making it explicit explicit to itself. 'Hegel' is that part of spirit (cultural software) which articulates the character of articulation itself. — plaque flag

Very nice :grin:

The reminds me of discussions of genesis versus structure. That concepts are open make genesis possible. As individuals we can get lucky with a new metaphor which gets adopted becomes relatively literal, hardens like cooling wax. Or we can add to the machinery of metacognition by seeing that maybe the inferential relationships of claims are what make concepts within such claims meaningful, etc. — plaque flag

Nicely put.

Jamal

The first is one I've expressed here often - many, I would say most, of the frustrating, fruitless discussions we have here on the forum start out with disagreements about the meaning of words and then never make any progress toward actually dealing with any interesting philosophical issues. — T Clark

It’s not clear to me whether this situation is the result of a lack of definitions, or an excessive focus on definitions. Perhaps you answer that when you go on to say…

I don't disagree that discussions where we work out among ourselves what particular terms mean are valuable. I have started a few discussions for that purpose - What does "mysticism" mean; What does "consciousness" mean; What does "real" mean. They were among the more satisfying discussions I've participated in. — T Clark

The unfolding of a concept in discussion :up:

On the other hand, I often start discussions about specific issues I want to examine, often something to do with metaphysics. In my OPs I often make it clear exactly what I intend the meaning of specific words are for the purposes of that particular discussion. Then I obnoxiously and legalistically defend that position, sometimes asking moderators to help. I do that because I want to talk about a specific concept or subject and I don't want to argue about what "metaphysics" really means. If I don't make those kinds of requirements, the thread will just turn into an argument about something I'm not interested in. — T Clark

I understand. This looks like stipulative definition, which I was mostly ignoring, treating it as something separate. Kant himself, though he says in those quoted passages that in philosophy you can’t start from definitions, clearly makes an exception in the case of stipulating how a term is to be used in his own work. E.g., “By synthesis, in the most general sense of the term, I mean the act of putting various presentations with one anotherl and of comprising their manifoldness in one cognition.” So the aim here is to be clear and open about a technical or provisionally restricted use of a term, because there is a particular argument you want to make.

Or maybe what you’re referring to is the exception in my main thesis, those times when a term is so ambiguous that you need to prevent confusion with a clear statement that this, not that, is what you mean.

fdrake

“Mathematical definitions never err,” he’s wrong? — Jamal

They're stipulated as true. But they can err in their expression of their intended concept. Like sets and unrestricted comprehension. In that case, what you mean by a collection of objects which satisfy a property should not also entail a contradiction when you use a property to pick out a set.

But although there cannot occur in the concept anything incorrect in content, sometimes–although only rarely–there may still be a defect in the form (the guise) of the concept, viz., as regards its precision — Kant, Critique of Pure Reason, B759

Think that about covers it. But I don't think it's that rare. If you've ever had to define a new structure, the definition itself is hard. You prove theorems about it, or can't prove theorems about it, and you jiggle around with the axioms until you can prove what you need to. That's very similar to provisional definitions and explications, then using them to refine the expression of the concept with a new definition.

There's an example of me pissing about like that in my old Value Theory thread.

Jamal

↪Manuel

:up:

A rose by any other name would smell as sweet. The killing of an old man, if such killing had a name, would be just as significant.

So, in general, I think that we most of the time, have a decent idea or notion of what we want to communicate. The failure of communication has more to do with the ideas behind the words, than the words themselves. So, I'm inclined to agree that philosophy shouldn't be primarily about definitions, though these can help. — Manuel

I agree, but I’m putting it more strongly: they can help, but they can also positively hinder.

Jamal

↪Fooloso4

Well yes, I made that point in the OP, and not only by quoting Kant. I also admitted that the title overstated the case. However, you did alert me to the fact that my central thesis also contradicted Kant, so thanks for that. Here’s the new version:

A definition of a philosophical concept might be required at the beginning of a discussion only in the case that the term is equivocal.

plaque flag

Examples are looked down upon by several philosophers, but they’re often what allow me to first get ahold of a concept, — Jamal

:up:

We see too that artificial intelligence learns from examples, which is probably mostly how we learn.

Jamal

Note the lower textual location in B only, this in reference to understanding, whereas the other quotes with higher textual locations, refer to pure reason’s dogmatic use, and is found in both editions.

Of course definitions have a place, if only in justifications for a method. — Mww

To argue that Kant believes that definitions have a place it’s clearer to just stick with the Transcendental Doctrine of Method, where he makes an exception for defining “concepts thought by choice” (invention), which I take to be stipulative definitions. And he uses these definitions at the beginning of certain sections of the CPR, e.g., “By synthesis, in the most general sense of the term, I mean the act of putting various presentations with one anotherl and of comprising their manifoldness in one cognition.”

Kant and I are setting aside stipulation as something quite different from the central issue here (though I’m not saying that this stipulation cannot be called a kind of definition).

plaque flag

↪Jamal

A little more from an inferentialist perspective:

If philosophers fail to define their terms, they get nowhere.

To evaluate this statement is to decide whether we (the royal we of universal rationality) ought to accept the inference as legitimate. To make a case for such a policy, we will have to use still other inferences involving still other concepts that are tentatively accepted as sufficiently legitimate to do so. To me this suggests a huge web of concepts related inferentially with more or less confidence and familiarity.

How is deciding the meaning of a concept like define related to deciding such legitimacy ?

Jamal

How is deciding the meaning of a concept like define related to deciding such legitimacy ? — plaque flag

It's the same thing?

plaque flag

It's the same thing? — Jamal

I think (?) inferentialism would say yes, but of course we have to think of all possible inferences involving 'define.' Personally I find this plausible. What after all do we do with concepts ? But decide what to do, often together ? And justify why we did it. Justification seems deeply and maybe essentially inferential.

This is a bit like understanding the bishop by talking about checkmating the king.

jgill

People who get stuck on specific definitions are often irritating pedants and seem to miss the point. — Tom Storm

Guilty. Just my math background showing. :yikes: And the old adage from CS: "garbage in = garbage out".

But, as @fdrake explained, definitions in math sometimes undergo revisions as the process of exploration or problem-solving progresses. Research in math is a very fluid work space if one is not forced to make it advantageous in some applied problem. It can be remarkably loose, going back and forth. I'm looking at something now that will necessitate a revision of hypothesis - the starting points, like definitions, for successful arguments.

Although it irritates me at times to read sloppy, ill- formulated definitions in these philosophical discussions, I am growing to understand it's part of the process that might converge to an interesting conclusion.

Jamal

I think (?) inferentialism would say yes, but of course we have to think of all possible inferences involving 'define.' — plaque flag

Is that what inferentialism entails? That's a bummer.

You're in danger of forcing me to read Brandom.

Tom Storm

Guilty. Just my math background showing. — jgill

I hadn't noticed. Digression. Do you think there is a math brain or a type of person to whom math speaks? I ask simply because I can't do any math at all. I am borderline innumerate.

Jamal

↪Tom Storm

I hope you don't mind if I jump in here, even though I'm not a mathematician and haven't really done any mathematics for thirty years.

My view is that there might be such a thing as a non-math brain, but it's rarer than people think. It's a bit like music: lots of people say "I'm tone deaf" but in my experience with some coaxing and a little light training they can learn to sing in tune, identify intervals and so on. In both cases it goes back to a bad education.

I did mathematics in my engineering degree but haven't used it since, hence I can't do it any more. Like music, it demands constant practice to stay on the horse, and without that it becomes very difficult to get back on.

I guess these are common observations or even platitudes, but I think they're importantly true.

One thing I can't do well is games, like chess and poker. I don't think there's something about my brain that caused me to turn out like this, more like a psychological thing, a neurosis or whatever. Similarly, telling oneself and others that one is borderline innumerate might just reinforce a psychological block that stands in the way of your mathematical genius.

Janus

A definition is a statement that specifies the correct use of a term. — Jamal

I disagree with this definition of definition, or at least I don't agree that it is the one correct definition or that there are correct definitions at all. I would say instead that a definition is a statement that specifies an interpretation of the meaning of a term. And that is why definitions are needed in philosophy, because otherwise interlocutors will waste time and energy talking past one another.

Tom Storm

I hope you don't mind if I jump in here, — Jamal

Not at all.

One thing I can't do well is games, like chess and poker. — Jamal

Now that's interesting, I don't do games at all. But I have no interest in them. I played a few boardgames as a kid and that's it. No cards, video games, nothing for 40 years.

telling oneself and others that one is borderline innumerate might just reinforce a psychological block that stands in the way of your mathematical genius. — Jamal

Ha! It would be nice to contemplate the possibilities, but alas I'll never find out.

Like music, it demands constant practice to stay on the horse, and without that it becomes very difficult to get back on. — Jamal

That's interesting. A form of math fitness, maybe. I hadn't considered that.

Definitions have no place in philosophy

Welcome to The Philosophy Forum!

Categories

More Discussions